A virus is an infectious agent too small to be seen directly with a light microscope. They are not made of cells and can only replicate inside the cells of another organism (the viruses’ host). Viruses infect all types of organisms, from animals and plants to bacteria and archaea. Viruses are found in almost every ecosystem on Earth and these minute structures are the most abundant type of biological entity. Viruses consist of two or three parts: all viruses have genes made from either DNA or RNA, long molecules that carry genetic information; all have a protein coat that protects these genes; and some have an envelope of fat that surrounds them when they are outside a cell. Viruses vary from simple helical and icosahedra shapes, to more complex structures. Most viruses are about one hundred times smaller than an average bacterium. Viruses spread in many ways; plant viruses are often transmitted from plant to plant by insects that feed on sap, such as aphids, while animal viruses can be carried by blood-sucking insects. Influenza viruses are spread by coughing and sneezing. Rotavirus is a genus of double-stranded RNA virus in the family Reoviridae. It is the leading single cause of severe diarrhea among infants and young children, and is one of several viruses that cause infections commonly known as stomach flu, despite having no relation to influenza. By the age of five, nearly every child in the world has been infected with rotavirus at least once. However, with each infection, immunity develops, subsequent infections are less severe, and adults are rarely affected. There are seven species of this virus, referred to as A, B, C, D, E, F and G. Humans are primarily infected by species A, B and C, most commonly by species A. All seven species cause disease in other animals [4].

Within rotavirus-A, there are different strains, called serotypes. As with influenza virus, a dual classification system is used, which is based on two structural proteins on the surface of the virion. The glycoprotein VP7 defines G-types and the protease sensitive protein VP4 defines P-types. Strains are generally designated by their G serotype specificities (eg, serotypes G1 to G4 and G9), and the P-type is indicated by a number and a letter for the P-serotype and by a number in square brackets for the corresponding P-genotype.(P-serotypes are difficult to characterize; therefore, molecular methods based on sequence analysis are often used to define the corresponding P-genotype instead. These genotypes correlate well with known P-serotypes). Because the two genes that determine G-types and P-types can be passed on separately to offspring, various combinations occur in any one strain [5]. Segmentation of cell regions is an important step in computer-aided analysis of rotavirus particle images in microbiology. Accurate and reliable segmentation is an essential step in determining valuable quantitative information on size, shape and texture, which may assist microbiologists in their diagnoses. Parametric algorithms were recently introduced for segmentation of cell images with elliptically shaped cells or cells whose contours were relatively smooth. The snakes or active contours are used extensively in computer vision and image processing applications, particularly to locate object boundaries. Nonparametric algorithms can generally be categorized as region-based, edge-based, histogram-based, clustering and neural network based algorithms [1]. For original noiseless cell images, most algorithms work fine. However, when an image is corrupted by heavy noise, these algorithms may not produce satisfactory segmentation results. The algorithm introduced in this paper deals with